CONSTANCE: SYSTEMATICS OF THE ANCIOSPERMS ' ^^y 



Little has as yet been i)ublished with regard to the phloem of dicotyledons, but 

 the pioneer work of Hemenway (1913) appears to indicate that the ])hyloge- 

 netic development of the sieve tube may to some extent parallel that of the vessel 

 element. The present status of our "meager and fragmentary" knowledge of 

 evolutionary trends of specialization in the phloem has recently been summar- 

 ized by Esau, Cheadle, and Gifford (1953). From a study of the monocotyledons 

 it appears that sieve tubes have undergone a progressive localization of sieve 

 areas on the end walls, a gradual shift from oblique to transverse end walls, 

 a change from compound to simple sieve plates, and a progressive decrease in the 

 prominence of sieve areas on the side walls. The sequence of development of 

 sieve tubes in dicots has yet to be established. In an extensive series of investi- 

 gations on monocots, Cheadle and his co-worker (Cheadle, 1937, 1938, 1939, 

 1942a, 1942b, 1943a, 1943b, 1944, 1948; Cheadle and Uhl, 1948; Cheadle and 

 Whitford, 1941) have established apparently phylogenetic trends in the form 

 and distribution of vessels and sieve tubes, and in the types of vascular bundles. 

 Most important, perhaps, is the discovery that in monocotyledons vessels de- 

 velop first in the roots and spread thence to the aerial parts; in contrast, the 

 sieve tubes are believed to have become modified in the aerial parts, and 

 to have progressed basipetally toward the root. 



Wood anatomists have been, in the main, extremely modest in making claims 

 as to the evolutionary significance of their data, and as to its utility for the 

 problems of the taxonomist. It has been emphasized repeatedly (Metcalfe, 1944, 

 1946; Bailey and Howard, 1941; Tippo, 1946; Bailey, 1949, 1951) that appar- 

 ently close affinity in structure may be due to parallel or convergent evolution 

 rather than to any close genetic relationship, and that anatomical data are more 

 valuable in indicating that a proposed phylogenetic connection is impossible 

 rather than that it is probable (Bailey, 1944b; Swamy and Bailey, 1949). Thus, 

 the anatomical method is to be regarded as an auxiliary one (Fritsch, 1903; 

 Solereder, 1908; Eecord, 1934; Vestal, 1937, 1940; Cheadle, 1942a; Chalk, 1944; 

 Metcalfe, 1946), but one which has the advantage that its postulated trends of 

 phylogenetic specialization have been propounded quite independently of pre- 

 conceived ideas as to the primitiveness of any particular group of angiosperm, 

 or of their derivation from any particular type of ancestor (Tippo, 1938, 1946; 

 Vestal, 1940). In the words of Bailey, "If a truly natural classification is to be 

 attained, it must be based upon the analysis and the harmonization of evidence 

 from all organs, tissues and parts" (1949, p. 66). 



Other Vegetative Characters 



The existence of conflicting theories, bolstered by little convincing substan- 

 tiating evidence, characterizes the analysis of phyletic trends in regard to 

 the vegetative structures of angiosperms. That the stems of primitive flowering 

 plants were unbranched or little branched, and that derived types exhibit richer 

 ramification, has been postulated by Arber and Parkin (1907), Bessej^ Bews 

 (1927), and Corner (1949). Nodal and petiolar anatomy were regarded by Sin- 

 nott (1914) as being essentially conservative. He suggested that the trilacunar 

 condition might be primitive and both the multilacunar and the unilacunar 

 states derived, as indicated by the persistence of a trilacunar arrangement in 



